Few Customers Seem to Buy "Best" Services Available

Supply and demand both matter when evaluating consumer internet bandwidth, even if most of the attention, praise or criticism centers on supply, not demand. Consider that, in Germany, half of mobile customers still buy 3G. What really matters is that consumers can buy service that matches their needs, not any arbitrary measure of supply. 

In every market, it seems, the percentage of customers who buy the fastest tier of service--typically the most expensive as well--is a small percentage of total buyers. In the U.K. market, about eight percent of buyers choose the fastest tier of service, at speeds greater than 120 Mbps. Altogether, 14 percent of U.K. customers buy any service operating at 100 Mbps or faster. 

source: Ovum

Still, 42 percent buy the slowest tiers of service, with 21 percent paying for 10 Mbps or slower service, while another 21 percent buy speeds between 10 Mbps and 20 Mbps.

In the U.S. market, where broadband is defined as a minimum of 25 Mbps, 42 percent of U.K. customers are not buying “broadband” service.

The point is that making higher speeds available is one thing. Getting customers to buy those services is quite something else. 

And yet headline speeds keep racing higher. Keep in mind that bandwidth supply for lead users in urban areas grows at Moore’s Law rates. More specifically, bandwidth for lead users has increased at rates consistent with Nielsen's Law, which predicts that internet speeds grow by 50 percent per year. 

 

source: Nielsen Norman Group

Perhaps improbably, U.S. cable company Comcast, for example, has been doubling the top speeds on its networks at rates consistent with Moore’s Law . “Comcast has increased speeds 17 times in 17 years and has doubled the capacity of its broadband network every 18 to 24 months,” Comcast says.

By doubling its capacity every 18 months, Comcast and other U.S. cable operators have achieved what most of us probably doubted would be possible.

Overprovisioning of bandwidth is not typically a problem, but supply still must bear some reasonable match to demand. Right now, consumers can do virtually anything on about 25 Mbps per concurrent user. That target of course will continue to grow over time, but networks really need to be able to supply about that much bandwidth anywhere, to be generally useful in the U.S. market. 

Headline speeds are one thing; consumer demand quite another.